Climate change-induced weather extremes are increasing the intensity and frequency of disturbance events, posing a major threat to forests globally. In Central Europe, hotter and drier summers, such as those during the 2018–2020 drought period, have led to widespread forest damage. To adapt forests to a hotter and drier future it is important to identify sites more vulnerable to canopy mortality during drought, but high-resolution tree mortality data across a continuous landscape is still sparse.</p> <p>This study aimed at filling this research gap by utilising a high-resolution (single-tree-level), spatially continuous dataset covering the entire Central European country of Luxembourg. We used generalized additive models (GAMs) to explore the contribution of biotic and abiotic site characteristics to the observed canopy mortality of conifer and broadleaf forests during the 2018–2020 summer droughts.</p> <p>Our model explained 44.6 % of canopy mortality variation in conifers and 25.3 % in broadleaf forests. Clustered mortality patches spreading from one year to the other, typical for bark beetle infestation, were the strongest predictor of canopy mortality in conifer trees. Forest height also emerged as a strong predictor of mortality in both forest types. Surprisingly, we found limited influence of topography on canopy mortality. Our study highlights the potential of using high-resolution canopy mortality data across a national-scale study area to unravel the influence of site characteristics driving spatial variation in forest mortality during drought events.